Piloting the National Cardiovascular Research Infrastructure: The SAFE-PCI for Women Trial
Sunil V. Rao MD Associate Professor of Medicine Duke University Medical Center Duke Clinical Research Institute
Disclaimer
The views and opinions expressed in this presentation are those of the individual presenter and do not necessarily reflect the views of the Clinical Trials Transformation Initiative.
SAFE-PCI for Women
Background and Rationale NCRI construct
Advantages Challenges
Trial results Lessons Learned
SAFE-PCI for Women
Background and Rationale NCRI construct
Advantages Challenges
Trial results Lessons Learned
TREATT I ThinkTank
The rate of radial approach is lower in the US compared with other countries
Lack of education and perhaps lack of large US-based randomized data may be responsible
Large appetite for a randomized trial looking at clinical outcomes
Challenge #1 is randomization Femoralists unable to randomize to radial
Radialists unwilling to randomize to femoral
Challenge #2 is funding
Post-PCI Bleeding and Vascular complications
Verheugt F, JACC Intv 2011
1-year Mortality
Kugelmass A, AJC 2006 Alexander K, et. al. Circ 2006
Overall
Women
1.46 (1.22, 1.73)
1.72 (1.30, 2.28)
0.5 1.0 1.5 2.0 2.5
Higher
Bleeding Risk
Lower
Radial approach
Bertrand OF, et. al. AHJ 2012 Feldman DN, et. al. Circ 2013
0
2
4
6
8
10
12
14
2007 -‐
2007 -‐
2008 -‐
2008 -‐
2009 -‐
2009 -‐
2010 -‐
2010 -‐
2011 -‐
2011 -‐
% Rad
ial
Females Males
Radial approach in men vs. women From the NCDR CathPCI Registry®
• Women significantly underrepresented in prior trials • Women present a unique challenge
• Higher bleeding risk but radial approach underused • Smaller radial arteries • Potentially higher transradial procedure failure rate
SAFE-PCI for Women
Background and Rationale NCRI construct
Advantages Challenges
Trial results Lessons Learned
Study of Access site For Enhancing PCI for Women (SAFE-PCI for Women)*
Female patient undergoing urgent or elective PCI
Best background medical therapy
Bivalirudin, Clopidogrel, Prasugrel
2b3a at investigator’s discretion
Radial Femoral N=1800 pts, 30 sites
Sites from NCRI
Patent hemostasis required
Vascular closure devices allowed
Primary Efficacy Endpoint: BARC Types 2, 3, or 5 bleeding or Vascular Complications requiring surgical intervention
Primary Feasibility Endpoint: Procedural failure
Secondary endpoints: Procedure duration, total radiation dose, total contrast volume
*Planned in collaboration with ACC, CSRC, FDA Office of Women’s Health
Methods – The National Cardiovascular Research Infrastructure
• Embeds randomization into the NCDR CathPCI Registry®
• Mechanism for identifying appropriate trial sites
• Leverages the workflow of registry participants by
electronically exporting trial-relevant data into an electronic
case report form
– Reduction of redundant data entry (~60% data needed for study
patients from CathPCI registry)
– Reduced trial costs due to reduced site-level workload
• Data output using CDISC SDTM standards
• 21 CFR 11 compliant – IND and IDE applications
Site identification Using actuaI data rather than PI recall!
Sites
NCDR PCI records from 2009Q3 through Jan 2011
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90
Male Female
Rad
ial P
CI V
olum
e
Methods - SAFE-PCI for Women workflow
Demographics Medical Hx
Procedural data
Autopopulate
Analytic Database
Unique pages for trial
Randomization
NCRI - Advantages
• Streamlines data collection/entry • Encourages collaboration between multiple
stakeholders at the site level – Research coordinators – Registry coordinators – Site Pis – Quality managers
• Minimizes costs by reducing site payments • Costs are generally up front – creation of the
software interface
NCRI Disadvantages
• Specific to the clinical trial data platform (e.g. InForm)
• Registry often is disease state specific – Data input may be automated and not conform to
clinical trial schedule – Fees for change may be high and not accounted for in
study budget • Multiple stakeholders = multiple priorities • Collaboration can be challenging at the site
level
SAFE-PCI for Women
Background and Rationale NCRI construct
Advantages Challenges
Trial results Lessons Learned
Methods – Patient population Inclusion
• Age > 18 years
• Female patient undergoing elective or urgent PCI or
• Undergoing diagnostic angiography to evaluate ischemic symptoms with the possibility of PCI
• Have capacity to sign informed consent
Exclusion
• Conditions precluding safe arterial access
– Non-palpable radial or femoral pulses – Bilateral abnormal Barbeau tests – Hemodialysis AV fistula or graft in arm
to be used for arterial access – INR ≥ 1.5 if on warfarin
• Bilateral IMA grafts • Planned staged PCI within 30d of
index PCI • Valvular heart disease requiring
surgery • Planned RHC • Primary PCI for STEMI
Two cohorts specified: • Total randomized – all women who are randomized regardless of whether they
undergo PCI • PCI cohort (primary analysis cohort) – Guidewire exiting the guide catheter for
diagnosis or treatment and therapeutic anticoagulation given
Methods - Endpoint definitions
Primary efficacy endpoint
• BARC Bleeding – Type 2: Overt, actionable bleeding not
meeting criteria for type 3, 4, or 5 bleeding
– Type 3: • Overt bleeding with hgb drop ≥ 3
g/dL (corrected for transfusion) • Transfusion with overt bleeding • cardiac tamponade • bleeding requiring surgical
intervention or intravenous vasoactive drugs
• intraocular bleeding or ICH – Type 5: Fatal bleeding
• Vascular complications requiring intervention
– AV fistula
– Pseudoaneurysm
– Arterial access site occlusion
• Access site crossover – Inability to complete the procedure
from the assigned access site
Primary Feasibility Endpoint
CEC Adjudication of all suspected bleeding or vascular complication
events
Methods
• Sample size calculation – Rate of BARC-type bleeding in NCDR CathPCI Registry among women
without STEMI ~ 8.7%1
– Assumptions • Femoral access bleeding or vascular complication rate – 8% • 50% reduction with radial access; 1576 patients provides 90% power at
alpha 0.05 • Sample size increased to 1800 due to uncertainty around event rates • 3000 women randomized to obtain 1800 women undergoing PCI
• All primary analyses performed by modified intention-to-treat • Primary analysis in PCI cohort; Sensitivity analysis in Total
Randomized Cohort • Three subgroups examined for primary efficacy endpoint
– Prespecified in PCI cohort: ACS vs. non-ACS, Site radial volume – Post-hoc in Total Randomized Cohort: PCI vs. no PCI
1Rao SV, et. al. JACC Intv 2013
Trial conduct
• After 1120 women had been randomized, routine review of trial endpoints by DSMB – Primary efficacy event rate markedly lower than expected
– Trial unlikely to show a difference at the planned sample size
– Recommended termination of the trial
• No harm noted in either the radial or femoral groups
• Steering committee voted to continue study until enrollment in a quality-of-life substudy was complete (N=300)
Results - Final Recruitment
96.7% of sites enrolled ≥ 1 patient 70.9% of sites enrolled ≥ 10 patients
1787 women randomized At 60 US sites
893 women assigned to Radial 894 women assigned to Femoral
Secondary 30-day endpoints 290 PCI pts 292 PCI pts
ITT: Primary 72 hr or discharge endpoints
891 women 345 underwent PCI
884 women 345 underwent PCI
Results – Primary efficacy and feasibility endpoints PCI cohort
• Interactions for primary efficacy endpoint not significant for ACS vs. Non-ACS, tertiles of site radial volume
• Most common reason for needing to convert from radial to femoral access to complete the procedure was radial artery spasm (42.9% of crossovers)
Radial (N=345)
Femoral (N=346)
OR (95% CI)
P
BARC 2, 3, 5 bleeding or Vasc Complications
1.2% 2.9% 0.4 (0.1-1.3) 0.12
Access site crossover
6.1% 1.7% 3.6 (1.5-9.2) 0.006
Results – Primary efficacy and feasibility endpoints Total randomized cohort
• Interaction term for primary efficacy endpoint not significant for PCI vs. no PCI • Most common reason for needing to convert from radial to femoral access to
complete the procedure was radial artery spasm (43.6% of crossovers) • Only one patient did not have the procedure successfully completed – was
randomized to femoral
Radial (N=893)
Femoral (N=894)
OR (95% CI)
P
BARC 2, 3, 5 bleeding or Vasc Complications
0.6% 1.7% 0.3 (0.1-0.9) 0.03
Access site crossover
6.7% 1.9% 3.7 (2.1-6.4) <0.001
SAFE-PCI for Women
Background and Rationale NCRI construct
Advantages Challenges
Trial results Lessons Learned
Lessons learned
Get all parties involved EARLY Budget for the unexpected Consider the different missions of the registry vs.
the trial SAFE-PCI for Women – FFR example
Control the “trialist urges” and streamline the data collection
Don’t overestimate the cost savings Comes from reduced site work/payments Adjudication may be needed Core labs may be needed